The first-pass metabolism of 5-fluorouracil has been investigated in rats to compare systemic bioavailability after administration by different routes, the bioavailability after intravenous bolus administration being defined as unity. Bioavailability after oral administration (F(po)) was compared with that after intraintestinal administration into the closed loop (F(loop)) in conscious rats. F(po) was very low and variable (0.28 +/- 0.30, mean +/- s.d.), in agreement with earlier studies in man, but comparable with F(loop) (0.33 +/- 0.05), suggesting insignificant loss of 5-fluorouracil by degradation in the gastrointestinal lumen or by faecal excretion. The bioavailability after intraportal vein administration (F(ipv)) was compared with F(loop) in rats anaesthetized with pentobarbital, anaesthesia being used to maintain a stable portal drug infusion that mimics the sustained input of drug into the portal blood flow after intra-intestinal administration. F(ipv) was smaller than unity (0.68 +/- 0.03), suggesting significant hepatic first-pass metabolism, but higher than F(loop) (0.31 +/- 0.10), suggesting significant first-pass metabolism in the intestinal mucosa. The intestinal bioavailability for passage through the epithelial mucosa (Fi) was estimated, from the ratio of F(loop) to F(ipv) to be 0.46. The study revealed that both the liver and intestinal mucosa are responsible for the extensive first-pass metabolism of 5-fluorouracil after oral administration. This first-pass metabolism might be similar to that in man, in which the oral bioavailability is reportedly similar to that in the rats used in this study. The findings in this study should be of help in monitoring ways of improving oral 5-fluorouracil therapy.